So in the Musk thread you'll stop criticizing the cars themselves but instead begin hoping they succeed.
I am far from convinced Tesla's will save the world. I think they take too much rescorces to be a true mass transport option. They may end up as a way for the rich to insulate themselves from the rest of the world.
I actually just think software developers are too OK with failure and too uncritical of, well, critical requirements
I kind of agree, but it is not really the software engineers fault.
The low standard of software in critical roles really irritates me, but I put it down to the general attitude of "We want this, what is the safest/cheapest way to make it happen" rather than "what can we do safely".
The primary way this shows is the amount of personal data that is lost by organisations that could do better, with recent examples of organisations that one would have thought WOULD care about their staffs data not being spread around the dark web would be the UK
MOD and
police who have both lost large amounts of staff data in recent months.
The police payroll system seems a perfect example. The police used to have a payroll system that was not on the internet. At some point someone decided to change to an internet connected system, knowing that this data breach was at least a possibility if not a likelihood or even certainty given enough time. They went ahead anyway. This decision was made by organisations all over the world, and now most (?) peoples data is in the hands of hackers.
To bring it back to cars, the
range rover wireless key thing was a case where they had a secureish system and they made it loads worse because they wanted the feature that people could unlock their cars without doing anything (?). Their solution to the obvious security hole is not to design a secure system (perhaps a switch on the key) but to
ban rf equipment, which is only becoming more important in our lives.
Basically, a piece of software that becomes big enough and touches millions of people needs to be tweaked and worked on on a regular basis to adjust for the ever shifting reality, to adopt good ideas from competition. More importantly: to remove flaws, irrelevant code. That's just the nature of the thing. Otherwise: decay, stagnation, death.
And then we come to software updates. Because of the standards that they hold themselves to they "need" OTA updates. They creates the obvious security hole that now your car can be hacked wirelessly. Are you confident that these people, who cannot design a key that is secure at release, can design a OTA update system that will be secure for the lifetime of the car? If it is not, what is the likely cost when "how to hack a car and make the brakes fail at speed" is a popular video on tiktok?
They could have just made it require a physical connection, right? Or required it to be taken to a garage? There were secure options, but they were rejected for convenience.
3. Sometimes new apps come out, games, that need integration. The CIV clone "Polytopia" was a big hit among Tesla owners at one point.
As an example of another obvious way to solve this problem, I have bought a car without a stereo and with a large space where one was. I suspect the original ones will be very expensive, so I am thinking about just putting a tablet there, perhaps with a solid state amp. This will give me all the "multimedia" functionality of a fancy connected car without any of the security problems, and I will still be able to turn the heater up without taking my eyes off the road because I have real knobs.
Having a hard airgap between safety critical features of a car and the internet seems such a good thing and so easy to implement I am amazed it is not offered.
For obvious reasons, they don't let you drive and play simultaneously. It's either one or the other.
So your passenger cannot play games on the move? Elon is such a kill joy.